A Structural Model of Students’ Science Identity in the Context of GenAI

Source abstract

Abstract The application of GenAI in science education research has sparked debates on how students perceive themselves as science persons when they use GenAI to engage in scientific activities. Some studies suggested that GenAI fostered students’ interest in science, whereas others contended that GenAI might diminish students’ socialisation into disciplinary norms and science practices. In this study, we explored how students self-reported their science identity in relation to their use of GenAI in scientific activities. In this study, science identity in the context of GenAI is not a completely distinct construct from science identity. Instead, students can report different levels of science identity when they recall experiences of using GenAI in scientific activities. Using a survey design, this study developed and tested a structural equation model that examined how components of science identity interacted within GenAI-mediated scientific learning. It contextualises the construct of science identity in the context of GenAI for two main attempts. First, this study examined the interplay among interest , recognition , and performance/competence , as well as the overall science identity of 239 Chinese undergraduate students in the context of GenAI. Second, it examined how four background variables of undergraduate students– frequency of using GenAI in scientific activities , socio-economic status , age , and sex – were associated with different components of science identity during students’ use of GenAI. Results indicated that only did recognition by others as a science person predicted those student participants’ overall science identity in the GenAI context. This suggested that developing students’ interest in science via GenAI cannot be assumed to lead to an ultimate outcome of overall science identity.